Google’s AI Now Cleared to Paint Pentagon’s Bathrooms, Maybe More
KEY POINTS
- •In January 2026, Google DeepMind's CEO Demis Hassabis said he's comfortable with working alongside the Pentagon after 2018’s internal backlash.
- •Google updated its AI principles in 2025, dropping restrictions on military and surveillance projects to pursue more government contracts.
- •DeepMind’s VP Tom Lue cited 'back office' and clerical uses for their Pentagon AI contracts, reassuring staff about ethical safeguards.
In January's Google DeepMind town hall—because what’s better entertainment than corporate AI ethics—CEO Demis Hassabis declared he’s 'very comfortable' with AI helping the Pentagon, finally conquering his 2018 employee revolt PTSD. VP Tom Lue proudly revealed Google’s 2025 AI policy rewrite: goodbye pesky pledge against weapons and surveillance, hello 'robust process' that lets AI turn Pentagon drone data into paper clips or whatever 'back office clerical work' entails. Meanwhile, Anthropic was getting banned and suing the DoD for drawing red lines nobody else noticed. Google’s cozying up to Uncle Sam, juggling biosecurity, cybersecurity, and robotic pals from Boston Dynamics who promise not to become terminators. This corporate tightrope act promises to be thrilling, if you like your existential crises served with an AI slurry of PR spin.
Share the Story
(1 of 3)Source: Businessinsider | Published: 3/19/2026 | Author: Hugh Langley